VECTOR WIREAI INTELLIGENCE
NVDA$1,847+3.2%MSFT$512+1.1%GOOGL$199-0.4%META$728+2.7%AMD$184-1.2%TSM$212+0.6%PLTR$98+4.1%AI IDX4,821+1.9%
PKT
SEEDRefresh Models Deals Regulatory Sources

DeepSeek-V3 vs Qwen3.5 35B A3B

DeepSeekvsAlibaba44 shared benchmarks1232 head-to-head
BenchmarkDeepSeek-V3Qwen3.5 35B A3B
AA Agentic Index1.644.1
AA Intelligence15.429.9
AA-LCR41.368.3
AA-Omniscience-37.6-48.1
arc_challenge95.395.4
Artificial Analysis Coding Index2337
C-Eval90.190.2
critpt00.9
gdpval020.3
GPQA Diamond68.484.5
GSM8K96.790.1
hellaswag88.985.6
HLE5.247.4
HMMT 202527.589.2
hmmt_feb_202529.289
humaneval92.166.5
IFBench4172.5
ifeval87.391.9
LiveCodeBench v646.974.6
longbench_v248.759
MBPP75.470.8
mmlu89.481.1
mmlu_prox70.581
mmlu_redux90.593.3
MMLU-Pro81.285.3
mmmlu79.485.2
multichallenge31.460
OJBench2436
OmniScience Accuracy25.420.1
OmniScience Non-Hallucination23.314.6
OpenBookQA95.444.2
scicode35.837.7
supergpqa53.763.4
SWE-bench Verified4270
TauBench V3 - Banking4.74.9
Terminal-Bench 2.116.940.8
Terminal-Bench Hard15.226.5
vectara_answer_rate97.599.8
vectara_avg_summary_length81.794.9
vectara_factual_consistency93.989.5
vectara_hallucination_rate6.110.5
winogrande84.979.2
τ²-Bench Telecom (AA run)47.189.2
τ³-Bench4.768.9

Best tracked score per model per benchmark (default configuration; source-attributed). ↓ marks lower-is-better metrics. Open either model for its full surface, provenance and pricing.